BACKGROUND. Current treatment for children with Polyarticular Juvenile Idiopathic Arthritis (Poly JIA) results in approximately 50% of the patients demonstrating clinically inactive disease (CID) while on treatment. Over 40% of the children with Poly JIA are treated with a TNF antagonist biologic that is often started 6 months (mos) of disease onset. Over 50% of those demonstrating CID for up to 12 mos will worsen significantly within 3-12 mos after treatment withdrawal. SPECIFIC AIMS. In children with Poly JIA who have demonstrated CID for 6 continuous mos on anti-TNF biologic therapy who then discontinue the anti-TNF therapy: Aim 1. Assess the ability of quantitation of the serum level of the protein complex of Myeloid Related Proteins S100A8 and S100A9 (S100A8/S100A9) to effectively and accurately predict who will demonstrate disease flare within 8 mos; Aim 2. Assess the ability of an in vivo cytokine capture assay (IVCCA) that quantitates the in vivo production of TNF in those patients on anti-TNF monoclonal antibody therapy (infliximab or adalimumab) to predict those patients who will demonstrate disease flare within 8 mos; Aim 3. Use whole-genome gene expression profiling to identify genes that are differentially expressed in peripheral blood cells after >6 mos of CID and at time of disease flare to better understand the underlying biology of these clinical states and possibly identify important targets for the maintenance and loss of clinical and biological inactivity. METHODOLOGY. In 5 pediatric rheumatology centers, 120 children with Poly JIA demonstrating CID while on anti-TNF therapy will be enrolled and followed for up to 14 mos. In those demonstrating CID for >6 continuous mos, the anti-TNF therapy will be stopped. The primary outcome variable (POV) is disease flare within the next 8 months using a validated definition of disease flare. The sensitivity, specificity, + and - predictive values, and area under the ROC curve of S100A8/S100A9 and IVCCA assay to predict the POV will be determined. IVCCA results will be compared to real-time PCR for TNF mRNA. Peripheral blood whole genome expression will be determined using Affymetrix chips after 6 mos of CID and at disease flare. Gene expression analyses will identify differences between patients in CID that do and do not flare after stopping anti-TNF therapy and also compare CID and disease flare. This study leverages ongoing participation of 3 of the 4 centers in a NIAMS funded project to identify gene expression in all JIA patients. SIGNIFICANCE. If successful, this study will result in a better understanding of the effect of TNF antagonist therapy on the biology of JIA and a safer and more cost effective approach to the use of these profoundly important new therapies.